File size: 2,107 Bytes
a2fa148 56bb571 58e4cd1 56bb571 0a12f9d c4f139f 473bad1 58e4cd1 56bb571 58e4cd1 56bb571 58e4cd1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 |
---
license: cc
---
A Textual Inversion Embedding for Stable Diffusion, trained on SD2.1
Version 2 of this embedding now allows you to achieve that loose digital paint style with much greater flexibility and fewer artifacts. I highly recommend it over V1.
Simply call on the keyword laxpeintV2 (or laxpeint for the original) as in, single solo skydiver by laxpeintV2, overhead view of English country, above a skydiver, excitement and adrenaline
<strong>Interested in generating your own embeddings? <a href="https://docs.google.com/document/d/1JvlM0phnok4pghVBAMsMq_-Z18_ip_GXvHYE0mITdFE/edit?usp=sharing" target="_blank">My Google doc walkthrough might help</a></strong>
One odd artifact (though I think it's a funny kind of blessing from the gods of the latent space) is that skydiver prompt and multiple variations all generated a horse. No horses were included in the training data, so where it came from? Who knows! But I love it.
Generally, laxpeintV2 lets you prompt directly and simply for what you want and you'll be quickly on the right track, whether that's a cute animal, anthropomorphic animal, fantasy scene, sci-fi scene, landscape, interior, or whatever you dream up. This embedding works with your vision and rarely requires you to struggle with its quirks.
 |